Chemical shielding of H₂O and HF encapsulated inside a C₆₀ cage

The influence of external electrostatic environments on encapsulated molecules is a significant research area. Endohedral fullerenes, like C₆₀, are ideal systems for this investigation due to their potential use in diverse applications, including nanoelectronics, qubits, and contrast agents. A key question is the extent to which C₆₀ acts as a Faraday cage, shielding the interior from external fields. 'Molecular surgery' techniques have enabled the study of larger encapsulated molecules, particularly H₂O@C₆₀, which exhibits ortho-para nuclear spin conversion detectable through various methods. This observation implies a degree of interaction between the encapsulated water and the external environment despite the shielding effect of the fullerene cage. HF@C₆₀, with its similar dipole moment to H₂O, serves as an important comparative system. Debate exists about the interaction between H₂O and the C₆₀ cage, with conflicting computational results regarding charge transfer and the tunability of the encapsulated H₂O's position and orientation via various external stimuli. This paper directly measures the location of H₂O and HF within C₆₀ cages to address these uncertainties.

Previous research on endohedral fullerenes has explored their diverse applications and properties. Studies on H₂O@C₆₀ have focused on its ortho-para nuclear spin conversion, observed through NMR, infrared spectroscopy, and terahertz spectroscopy, as well as through capacitance measurements of bulk crystals. The synthesis and characterization of HF@C₆₀ has provided a comparative system. However, there has been significant debate regarding the extent of interaction between the encapsulated H₂O and the C₆₀ cage, with some suggesting electrostatic isolation and others suggesting tunability of the molecule's position and orientation using external fields or forces. Computational studies have yielded conflicting results regarding charge transfer and the degree of shielding provided by the fullerene cage.

This study used a multi-pronged approach combining several experimental and computational techniques: 1. **Scanning Tunneling Microscopy (STM) and Non-Contact Atomic Force Microscopy (ncAFM):** These techniques were used to probe the local density of states of H₂O@C₆₀ on Cu(111) and assess whether the presence of water affected the fullerene's electronic structure. STM images revealed a mixture of bright and dark molecules due to surface reconstruction of the Cu(111) substrate. ncAFM provided additional information on molecular height variations. No differences were observed between empty and water-filled C₆₀ molecules, suggesting minimal electronic structure perturbation. 2. **Valence Band X-ray Photoemission Spectroscopy (VB-XPS):** VB-XPS measurements were performed at the Diamond Light Source to confirm the lack of electronic interactions between the encapsulated water and the C₆₀ cage. Spectra of a bulk film of H₂O@C₆₀ were compared to those of pure C₆₀, showing no discernible differences, confirming the lack of electronic interaction between the encapsulated molecule and the cage. 3. **Normal Incidence X-ray Standing Wave (NIXSW):** The NIXSW technique was employed to precisely determine the intra-cage position of both H₂O and HF molecules within C₆₀ adsorbed on Ag(111). The location of H₂O was probed via the O 1s core level, and HF via the F 1s core level, enabling precise determination of their positions relative to the substrate. LEED patterns confirmed the formation of a single-layer molecular superlattice. The experimental values for the coherent fraction and position were analysed using Argand diagrams to consider static disorder in the molecular positions. 4. **Density Functional Theory (DFT) and Molecular Dynamics (MD) Simulations:** DFT calculations, using VASP and CP2K packages with DFT-D3 dispersion correction, were performed to model the geometries of H₂O@C₆₀ and HF@C₆₀ on Ag(111) in both vacancy and atom-top adsorption sites. MD simulations at 180 K using CP2K were performed to investigate the internal rotational dynamics of H₂O@C₆₀. Results were compared to experimental NIXSW results. The data from all methods were analysed in conjunction to obtain a comprehensive view of the interaction between encapsulated molecules and their environment. The Argand diagram analysis was crucial in reconciling the experimental and theoretical results by including both static and dynamic disorder effects in the NIXSW interpretation.

The key findings of this study are: 1. **Lack of Electronic Interaction:** STM, ncAFM, and VB-XPS measurements consistently showed that the encapsulated H₂O and HF molecules do not contribute to the frontier orbitals of the C₆₀ cage, demonstrating a significant degree of chemical isolation. 2. **Off-Center Intra-cage Position:** NIXSW analysis revealed that both H₂O and HF molecules reside in off-center positions within their respective C₆₀ cages. This is attributed to the strong electrostatic fields induced by the fullerene's adsorption onto the metal substrate. 3. **Significant Internal Rotational Motion:** MD simulations indicated significant internal rotational motion of the encapsulated H₂O molecule, consistent with the NIXSW data and explaining the observed broadening of the O 1s core level peak. 4. **Mixed Adsorption Site Model:** A mixed adsorption site model, comprising both atom-top and vacancy adsorption sites for the fullerene, was found to best explain the experimental NIXSW data. Argand diagram analysis confirmed the consistency of experimental data with a mixed adsorption model where both adsorption sites contribute significantly, leading to considerable static disorder. 5. **Influence of Intra-Cage Electrostatic Field:** The differences between core level peak broadening and shifts for bulk and monolayer samples were attributed to changes in the intra-cage electrostatic environment resulting from fullerene adsorption. MD simulations provided semi-quantitative support for this effect. The DFT calculations provided structural information; however, direct comparison to experimental NIXSW results required consideration of the dynamic and static disorder introduced by the mixed adsorption model, highlighting the complexity of the adsorption process and its effect on the encapsulated molecules.

This research directly addresses the long-standing question of electrostatic screening and decoupling of fullerene-encapsulated molecules from their external environment. While the lack of electronic interaction between the encapsulated molecules and the C₆₀ cage is clearly demonstrated, the strong influence of adsorption-induced electrostatic fields on the molecules’ intra-cage positions is equally important. The off-center positions of H₂O and HF, along with their significant internal rotational motion, highlight that the fullerene cage, while chemically isolating the encapsulated molecules, does not completely screen them from the effects of surface interactions. The mixed adsorption model necessitates a reassessment of how the NIXSW data is analysed, revealing the complexities of the adsorption geometry and the importance of dynamic and static disorder in the system. This comprehensive study moves the understanding of fullerene encapsulation beyond simple models of electrostatic shielding, emphasizing the role of both chemical and electrostatic interactions.

This study demonstrates the chemical isolation of H₂O and HF within C₆₀ cages while simultaneously highlighting the significant influence of adsorption-induced electrostatic fields on their intra-cage positions and dynamics. The findings challenge the simplistic view of C₆₀ as a perfect Faraday cage, demonstrating the importance of surface interactions and the complexities of adsorption geometry. Further research could explore different substrate materials, fullerene sizes, and encapsulated molecules to establish a broader understanding of the factors influencing electrostatic shielding in these systems and investigate temperature-dependent dynamics of the encapsulated molecules in greater detail.

The study's interpretation of NIXSW data was complicated by the presence of mixed adsorption sites, leading to static disorder. The MD simulations, while informative, provide a semi-quantitative description of the observed effects, and further refinements in the computational models could improve the accuracy of predictions. The study primarily focused on Ag(111) as a substrate. Further studies using other substrates are needed to determine the substrate-specific influence on the electrostatic environment within the fullerene cage.